This project involves a study of the recombination events that cause chromosome rearrangements and are useful as assays of recombination. In these assays (unlike sexual crosses) the recombining sequences include no provided structures likely to initiate recombination (such as double strand ends). Thus exchange events are being initiated by spontaneously occurring structures (nicks, breaks, displaced strands) that can be recognized by any recombination pathway. Pathways that can make full exchanges (essential for inversion) can be distinguished from pathways that perform only half exchanges (sufficient for duplication formation). The use of short sequences, inserted at various points around the chromosome, allows us to detect the influence of neighboring sequence differences on recombination. The phenomenon of 'forbidden' inversions has been attributed to a paucity of initiating structures near the recombining sequences. This model will be tested by isolating mutants that allow inversion of 'forbidden' intervals and by inducing chromosome breaks near the recombining sequences to initiate recombination. Some mutants are already in hand that permit inversion of 'forbidden' intervals. Available evidence for reciprocal exchanges in bacteria is less than secure. Using a 'twin spot' colony assay, to look for the association of duplication and deletion events, it should be possible to demonstrate reciprocality of exchanges performed by wild type cells and by various rec mutants. Transduction assays will be used to test particular aspects of recombination. A 'short sequence' assay reveals a contribution of recJ, recF and sbcB proteins to recBC-dependent recombination. This assay requires high speed recombination because the event must be completed before ends are degraded. Thus the assay may reveal accessory functions that contribute to the speed of recombination. We will use 'locked-in' P22-Mud prophages to generate homogeneous populations of transduced fragments for a particular region of the chromosome and will genetically determine the end point (or phasing) of these fragments. This should permit a genetic and physical analysis of the events that effect the end of a transduced fragment. Finally we will test a model to explain the phenomenon of 'adaptive mutation'. Point mutations are slightly leaky. When these mutants are placed on selective medium there is a strong selection for amplification which gives multiple copies of the mutant gene (per cell) and a small clone of such cells which are able to grow. Ultimately enough copies are formed to realize a correcting base substitution which allows rapid growth. Of the many copies of the gene, all subject to random mutagenesis, only the revertant copy is selected and detected by the assay. This gives the appearance of directing mutability to a few sites.